CODEWITHSUNDEEP
pythonpandasconditional-statements
N91
N91

Reputation: 415

Python Pandas multiple condition assignment

I have two pandas dataframes. One is the source and the other is the destination. I want to update based on multiple conditions the values of both dataframes. source dataframe look like this:

     Old_ID    New_ID   DATE      dt_insert
     FIRM345   FIRM21   21.11.19  11.11.19
     FIRM321   FIRM41   19.10.19  18.10.19

destination dataframe looks like this

     Old_ID    New_ID   DATE     
     FIRM345   FIRM21   21.11.19
     FIRM321   FIRM41   19.10.19

i want to know if there is a way to apply the following logic without using loops:

if src.old_ID == dest.old_id AND src.new_id == dest.new_id AND src.date == dest.date

THEN dest.dt_insert = src.date

ELSE append src row to destination dataframe

Upvotes: 1

Views: 103

Answers (2)

Eduard Ilyasov
Eduard Ilyasov

Reputation: 3308

You can solve your problem using this approach:

  1. outer join destination dataframe with a source dataframe on multiple keys (Old_ID, New_ID, DATE);
  2. replace a value in dt_insert column with a value from DATE column if the observation's merge keys are found in both dataframes;

  3. delete auxilary column _merge.

    import pandas as pd

src_data = [{'Old_ID': 'FIRM345', 'New_ID': 'FIRM21', 'DATE': '21.11.19', 'dt_insert': '11.11.19'},
            {'Old_ID': 'FIRM321', 'New_ID': 'FIRM41', 'DATE': '19.10.19', 'dt_insert': '18.10.19'},
            {'Old_ID': 'FIRM333', 'New_ID': 'FIRM31', 'DATE': '20.10.19', 'dt_insert': '20.10.19'}]

dest_data = [{'Old_ID': 'FIRM345', 'New_ID': 'FIRM21', 'DATE': '21.11.19'},
             {'Old_ID': 'FIRM321', 'New_ID': 'FIRM41', 'DATE': '19.10.19'}]

df_src = pd.DataFrame(src_data)
print(df_src)

#        DATE  New_ID   Old_ID dt_insert
# 0  21.11.19  FIRM21  FIRM345  11.11.19
# 1  19.10.19  FIRM41  FIRM321  18.10.19
# 2  20.10.19  FIRM31  FIRM333  20.10.19

df_dest = pd.DataFrame(dest_data)
print(df_dest)

#        DATE  New_ID   Old_ID
# 0  21.11.19  FIRM21  FIRM345
# 1  19.10.19  FIRM41  FIRM321

df_dest_new = pd.merge(left=df_dest, right=df_src, how='outer', 
                       on=['Old_ID', 'New_ID', 'DATE'], indicator=True)
df_dest_new['dt_insert'] = df_dest_new[['DATE', 'dt_insert', '_merge']].apply(lambda x: x[0] if x[2] == 'both' else x[1], axis=1)
df_dest_new = df_dest_new.drop(labels='_merge', axis=1)
print(df_dest_new)

#        DATE  New_ID   Old_ID dt_insert
# 0  21.11.19  FIRM21  FIRM345  21.11.19
# 1  19.10.19  FIRM41  FIRM321  19.10.19
# 2  20.10.19  FIRM31  FIRM333  20.10.19

Upvotes: 1

Dr. sans
Dr. sans

Reputation: 45

This should work

import pandas as pd

data = {'Old_ID':['FIRM345', 'FIRM321', 'FIRM11'], 'New_ID':['Firm21','FIRM41','FIRM42'],
        'DATE':['21.11.19', '19.10.19', '19.12.19'], 'dt_insert':['11.11.19','18.10.19','18.12.19']}
data2 = {'Old_ID':['FIRM345', 'FIRM321','FIRM12'], 'New_ID':['Firm21','FIRM41', 'FIRM43'],
        'DATE':['21.11.19', '19.10.19','19.12.19']}
src = pd.DataFrame(data)
dest = pd.DataFrame(data2)

print(src)
print(dest)

if src.Old_ID.any() == dest.Old_ID.any() and src.New_ID.any() == dest.New_ID.any() and\
    src.DATE.any() == dest.DATE.any():
    dest['dt_insert'] = src.DATE
else:
    src.append(dest)

print(src)
print(dest)

Upvotes: 1

Related Questions